About a month ago, House Democrats released bill text for their paid leave proposal. When it came out, I discussed the various flaws in the design, including that it has no minimum benefit level, has work history requirements that exclude a large minority of new parents, and provides two-parent families twice as much leave as one-parent families. These flaws are easy to solve and are not present in many paid leave systems in peer countries, but neither Democrats nor paid leave movement types have any interest in fixing them.
One other flaw I discussed earlier is that it gives money to employers who want to set up their own paid leave plans in lieu of the federal plan. This design is contrary to earlier Democratic paid leave proposals, including the one endorsed by Joe Biden during his presidential campaign, which provided for a completely public program. The new design essentially lays the groundwork for a heavily privatized paid leave system in which the federal government gives money to companies so that they can pay private insurance companies to administer paid leave benefits for their employees.
Insofar as this is basically just a paid leave version of our health care system, it has a ton of problems that are probably familiar to people interested in policy issues. But this paid leave bill is even dumber than our health care system. To understand why, we need to do a deep dive into how the subsidies for private paid leave plans are designed.
The Private Paid Leave Subsidies
Under the scheme, when an employer chooses to replace the federal program with a private paid leave insurance plan, they are entitled to a cash grant from the federal government equal to:
90 percent of the product of the projected national average cost per employee of providing paid family and medical leave benefits … multiplied by the number of employees (pro-rated for part-time employees) covered under the program for such calendar year.
The “projected national average cost per employee” is a dollar figure that the Secretary of Treasury will publish each year. The bill does not give a specific formula for deriving it. It just tells the Secretary of Treasury to figure it out.
Once this dollar figure is published, employers that are seeking public subsidies for their private paid leave plans will take the dollar figure, multiply it by their number of full-time-equivalent (FTE) employees, and then multiply that number by 0.9. The resulting figure will be the amount of the grant the Treasury gives them.
Put simply: every employer who takes advantage of this subsidy program will get a grant equal to 90 percent of the average cost of providing a paid leave benefit that is equal to the federal paid leave benefit that they are displacing. They get this grant regardless of what their actual cost of providing the benefit is, meaning that employers whose actual cost is less than 90 percent of the average cost will be able to pocket the difference.
A Lot of Employers Have Below-Average Paid Leave Costs
To illustrate the problem here, it helps to imagine three different employers: (1) a low wage employer (LWE) that pays all of their employees $20,000 per year, (2) a middle wage employer (MWE) that pays all of their employees $50,000 per year, and (3) a high wage employer (HWE) that pays all of their employees $100,000 per year.
According to the wage-replacement formula in the bill, the weekly paid leave benefit is $317 for the LWE, $690 for the MWE, and $1,057 for the HWE.
The “projected national average cost per employee” of a paid leave plan should be set somewhere near the costs associated with the MWE’s plan. So, for the sake of illustration, let’s say that the MWE’s costs are equal to 100% of the projected national average cost per employee. This means that, holding all else equal, the LWE’s costs are equal to 46% (317/690) of the projected national average cost per employee and the HWE’s costs are equal to 153% (1057/690) of the projected average national cost per employee.
Remember from above that, despite these very different underlying costs, each employer is eligible for 90% of the projected national average cost per employee. This means that an LWE could provide a basic paid leave plan equivalent to the federal plan, apply for a grant from the Treasury, and then pocket about half of that grant (90-46).
What will almost certainly happen if this structure is passed into law is that insurance companies like Sun Life will start contacting companies whose projected paid leave costs are well below 90% of the national average and explain to them that if they enter into a deal with Sun Life, they can get free money from the federal government, and to sweeten the pot, Sun Life will even fill out the paperwork for them.
On the other end of the wage scale, it would be stupid for an HWE to enroll in the system because the grant they are eligible for is worth less than 60 percent of their costs (90/153). Thus, a smart HWE would simply not enroll in the program and have their workers receive the federal benefit instead. If the HWE wanted to provide supplemental benefits on top of the federal benefit, they could still do that more cheaply by setting up an unsubsidized supplemental plan rather than displacing the federal benefit.
Another way to put all this is that giving every employer a cash grant that is pegged to the average cost of paid leave massively incentivizes adverse selection into the program by lower-than-average-cost employers and and out of the program by higher-than-average-cost employers. This will waste tremendous amounts of money.
Medicare Advantage for Dummies
There is actually a federal benefit already on the books that allows individuals to opt out of the federal plan and into a private plan that then receives a subsidy from the federal government. It’s called Medicare Advantage.
When a person opts for a Medicare Advantage plan over Traditional Medicare, CMS pays a subsidy to the insurance company that operates their Medicare Advantage plan. If Medicare Advantage was designed the same way that this paid leave proposal is designed, CMS would simply calculate the average cost of a Medicare enrollee and then set the subsidy equal to that amount (or equal to a very high percentage of that amount).
But doing this would create an obvious problem: not all Medicare beneficiaries receive an average amount of care each year. An 85-year-old person with cancer will be massively more costly in a given year than a 67-year-old with a clean bill of health. If CMS paid Medicare Advantage plans the exact same dollar amount for each person they enrolled, then those plans would be heavily incentivized to sign up healthier seniors who use a below-average amount of care and heavily disincentivized from signing up sicker seniors who use an above-average amount of care. This kind of cream-skimming would be hugely profitable for the Medicare Advantage plans and hugely wasteful for the federal government.
But CMS doesn’t do this. Instead, every single senior in the Medicare system is assigned a risk score every single year based on their demographic characteristics and recorded health status. The subsidies paid to Medicare Advantage plans are based on those risk scores. Put differently, the subsidy amount for each individual who opts for Medicare Advantage is risk-adjusted in order to prevent precisely the kind of adverse selection that the paid leave plan is blatantly inviting.
Don’t get me wrong. Medicare Advantage is also a ridiculous undertaking of the federal government. Compiling and updating 61 million risk scores every year is a monumental task that is done only to prop up a totally unnecessary private Medicare market. But if you are going to allow insurance companies to cannibalize federal social insurance programs like this, you need this kind of risk adjustment. The paid leave proposal does not do this and probably cannot do this.
If an employee of a paid leave non-profit organization came up with this design, then Melinda Gates needs to take them off the payroll immediately. If it was a lobbyist for the insurance industry, as is more likely, then their salary should be doubled.